Every inspection line eventually needs logic that no off-the-shelf tool provides. Building it has traditionally meant engineering time, and engineering availability. ReliVision 6.1 changes where that work happens.

Describe it instead of building it

With NLP & Image-Prompted Inspection Coding, users describe the inspection task in plain language and add a reference image. ReliVision generates a custom inspection step and places it directly in the pipeline. The reference image gives visual context for what the expected result should look like.

Refine instead of restart

When the first result is not right, users add further instructions and additional examples, and the step improves. Each iteration carries the context of the ones before it, so the reasoning behind a deployed inspection stays visible — including what changed, by whom, and when.

A practical starting point: weld X-ray porosity

One use is accelerating annotation. For example, in weld X-ray inspection, generated code can surface initial porosity candidates for review inside ReliVision turning a blank pipeline into a working detector and giving teams structured, labeled data for deep-learning training far sooner.

GenAI inside the inspection workflow

ReliVision 6.1 is not a coding assistant sitting beside the platform. It is generative AI inside the production inspection workflow, where the inspection actually runs.

Download ReliVision or contact info@relimetrics.com to request a demo.